| |
Abstract:
In a series of studies, we document the "eye movement-based
memory effect", showing significant changes in the way in which
subjects view visual materials as a function of previous exposure.
This effect (1) is seen for faces or scenes, across a range of
different instructional manipulations, (2) scales with the number
of prior exposures, (3) is manifested whether the previously viewed
items are subsequently presented in test displays individually or
intermixed among other (novel) items, and (4) can occur in the
absence of explicit remembering. Used as an indirect measure of
memory, our eye movement studies provide evidence for intact memory
in amnesia for the previous occurrence of faces or scenes, but not
for the relationships among objects in a scene. That is, amnesic
patients show the same repetition effect seen in normal control
subjects, i.e., changes in viewing patterns as a function of the
repetition of individual faces or scenes, but fail to show the
relational manipulation effect seen in normal control subjects,
i.e., changes in viewing due to manipulation of the relationships
among objects in previously studied scenes. These results document
a powerful new method for assessing memory, and they suggest that
the fundamental deficit in amnesia is an impairment in relational
memory binding (declarative memory).
|
